Never give up. The sky is the limit.

01Who I am

Every life counts. Not 999 out of 1,000.

Before I built software, I learned two professions in which there is no such thing as 99.9 %: airline pilot with Austrian Airlines and air traffic controller with Austro Control. The entrance tests for both are among the toughest in the world – the DLR test in Hamburg with a failure rate north of 98 %, the multi-day assessment at Austro Control in which people find out how resilient they really are. Tens of thousands apply, a few dozen get through. I passed both.

IQ measures how fast someone thinks. The cockpit test measures how calm someone stays when the display at the top right has been showing something wrong for five seconds and you have to be the first to notice. In aviation, the second one counts. And above everything stands one principle: every life counts. Not 999 out of 1,000. All of them.

A professional pose or an operating system? When I took up tennis again in autumn 2025 after a thirty-year break, I looked for a coach who had worked with ATP top-10 players for decades. Not because I want to go on tour – but because when I take something seriously, I want to know where my ceiling really is.

02At home

Three children, one father, no autopilot.

I am a single father of three children. The two younger ones are of school age, the eldest is an adult. Anyone who knows this knows: there is no roster that holds and no day that ends the way it was planned. You learn to set priorities in seconds – and to re-sort them the next morning.

My eldest son is severely disabled. He lives with a rare genetic defect, a microdeletion. He has taught me more about patience, reliability and what really matters than any assessment and any project.

My three children are the centre of my life and my motivation – every single day.

When I say “every life counts”, it is no longer a phrase from the cockpit. It is everyday life. And it is the reason I build software that gives people time back – time that is needed more urgently elsewhere.

My life has changed massively through all of this. I have become even more structured than I was used to from the cockpit – and that is exactly why, in the early morning hours before the house wakes up, there is room every day for an hour of sport: kettlebell, rope skipping, boxing. Not despite the responsibility, but because of it.

I am convinced: what matters most is the daily routine.

Big goals are not reached with big days, but with many ordinary ones that reliably start the same way. That is exactly how we build software: it should carry the daily routine – unspectacular, reliable, every day – so that head and time remain for the unplannable.

Never give up.
The sky is the limit.

My motto. Not a poster on the wall – a decision, every morning.

03One sentence in Hamburg

How KSIC became CloudDNA.

I founded CloudDNA more than ten years ago – originally not under that name. I came from SAP, where I had spent years at one of the largest software companies in the world. A safe harbour. I left it anyway.

It started with an unwieldy name: KSIC GmbH – Koch Security, Integration and Cloud Consulting. Four letters nobody could pronounce. Rescue came over dinner in Blankenese, Hamburg, at a fish restaurant on the Elbe. My old SAP colleague Jörg listened to my explanation for the third time, raised his glass and said a single sentence:

“Martin, you have the cloud in your DNA.”

Two weeks later KSIC GmbH was no longer called KSIC. Daniel, my first employee, was already there – on his very first business trip. (Daniel, I am still sorry about the turbot.)

We started in a tiny office in Oberwart. Today we work in the former Raiffeisen bank in Loipersdorf, in the middle of the spa and wine region of southern Burgenland, and draw on two talent schools you will not find on any world map: HTL Pinkafeld and the University of Applied Sciences Burgenland. The duo has become a team of 24 people, and we have published seventeen books with SAP PRESS – always from practice, never from marketing.

04The upheaval

Custom software was a red rag. Until 2023.

I built CloudDNA to create SAP integrations for large corporations – heavy software, paced in months. In that world custom software was avoided like the plague: too expensive, too slow, too risky.

Then came the language models.

I realised: this is not a gradual improvement, it is the upheaval that rewrites our craft. So I sat down myself and built software – at night, in my own editor, with the same tools I recommend to my team. You cannot explain AI-assisted development to anyone with a PowerPoint. You have to live it.

What used to take three months and a six-figure sum now takes weeks. Not because programmers work magic, but because copilots take over the drudgery and developers become what they should be: craftspeople with a head.

05We build it

From “We deliver” to “Wir bauen das”.

For more than ten years our promise was “We deliver”: a customer describes, we deliver. That made us big. But it has a limit: whoever only delivers what is ordered rarely builds what is needed.

That is why we are rebuilding ourselves. Consultants and developers are becoming forward deployed engineers and product owners. Forward deployed means: they sit where the problem arises – in the property management office, at the winery, in the law firm – and build on site. Product owner means: they stay until it is solved and own the product, not the ticket.

“Wir bauen das.” Three words you may only say to a customer if you mean them.

06We speak AI

Inside the AI bubble and, above all, outside it.

Inside the AI bubble everyone talks about models, benchmarks and agents. Outside – with the property manager running a thousand units, the winemaker in southern Burgenland, the restaurateur with twenty tables – a different question counts: Will this save me an hour tomorrow, and can I rely on it?

We translate between these worlds. We build with Claude by Anthropic every day and are well on the way to becoming an Anthropic partner. But what reaches the customer is not a model, it is a workflow that works:

• The AI proposes, a human decides. More autonomy comes only after a passed probation period.
• Every action is in the audit trail. What the AI did can be read back.
• AI has to pay for itself. Own key, fixed cost cap per tenant.
• Data stays where it belongs. Tenants separated down to the database.

Anthropic and Claude are trademarks of Anthropic PBC. The mention describes how we work.

07The Tailor family

Made to measure instead of standard with a hundred compromises.

Instead of just talking about AI-assisted development, we picked industries I know well enough – or in which I have friends who explain the problem to me until it sticks. Every Tailor shares the same foundation with its siblings, but speaks the language of its industry:

• ProbeTailor – for acceptance tests in SAP projects and business software, because in all my years of SAP consulting I have seen too many acceptance tests in Excel lists and too many screenshots in e-mails.
• JournalTailor – for accounts payable before SAP, BMD and DATEV, because in SAP projects I have too often watched invoices being retyped although the posting behind them has long been routine.
• DocumentTailor – for SAP shops, because I spent fifteen years building forms on SAP Adobe Document Services and did not want to leave the succession question unanswered any longer.
• RecordTailor – for everyone who needs a DMS that does not just store but reads – because I have too often seen someone secretly open Excel because the DMS is too slow.
• ImmoTailor – for property managers and developers, because accounting, maintenance and portals get in each other’s way there more than anywhere.
• TaxTailor – for tax firms, because a friend told me about his Windows terminal server over a glass of wine.
• DineTailor – for restaurants, because for years I have been sitting at other people’s tables selling champagne to restaurateurs.
• StayTailor – for hotels, because the spa region is full of businesses ducking under the booking portals’ terms.
• MeetingTailor – for meetings and boards, because foundation boards, councils and club committees among my friends keep describing the same late-night scene.
• Weintailor – for winemakers, because we sit in the middle of a wine region and every other trip turns into a wine trip anyway.
• BuildTailor – for construction and trades, because defects, plans and daily reports happen on site – not in the office.
• iFlowTailor – for SAP Cloud Integration, because I co-wrote the cookbook on it and still had ten tabs open every day: monitor, Postman, XPath tester, keystore, the Excel list of names. The workbench was missing, so we built it.

CloudDNA is therefore no longer just an SAP supplier, but a workshop with a growing product line. Bolder in business. Humanly the most honest thing we could do.

08Why iFlowTailor

I co-wrote the cookbook. And still had ten tabs open.

In a cockpit you have every instrument in front of you – not spread across ten windows. Integration development was never like that: editor here, monitor there, Postman, XPath tester, keystore, the Excel list of names. Six thoughts that iFlowTailor grew out of:

1. Thought 1 · The cookbook

   The knowledge was in the book. It was needed in the editor.

   Together with Thorsten Reisinger and Marc Urschick I wrote the SAP Cloud Integration Cookbook¹. Recipes for error handling, retries, security, performance. The catch: when you are building, the book is rarely lying open next to you. The 40 rules of Quick Review² are that knowledge, right where it is needed – and known error patterns point to the matching chapter.

2. Thought 2 · The cockpit

   All instruments in front of you, not in ten windows.

   On an aircraft nobody looks for the altimeter in another room. iFlowTailor is a panel next to the editor³ that knows which iFlow is open – review, test, monitor, keystore and search all refer to it.

3. Thought 3 · No server

   Data that isn’t stored doesn’t need guarding.

   Every other piece of Tailor software has a foundation with a database. iFlowTailor deliberately has none: no server, no telemetry⁴, just the existing SAP session⁵. That makes approval by IT shorter and the night quieter.

4. Thought 4 · Read-only

   Look first, then touch.

   A tool that can change a production tenant doesn’t have to do so out of the box. Write actions are off⁶ until someone deliberately switches them on – and then with before/after and a log. Like the checklist before starting the engines.

5. Thought 5 · Cause before symptom

   A hundred errors, one cause.

   In operations I spent too many nights opening failed messages one by one. Error analysis groups them by signature⁷ and knows the usual suspects: TLS handshake, 401 from the receiver, certificate chain, timeout.

6. Thought 6 · Hand on the lever

   The autopilot suggests. The human flies.

   Ask Tailor explains, drafts and advises – with your own model, masked⁸. Even the iFlow generated from a requirement is only created after confirmation and never deployed⁹. That isn’t caution born of distrust in the technology. It is the same rule I flew by.